Laser & Optoelectronics Progress, Volume. 57, Issue 12, 121020(2020)
Dunhuang Mural Inpainting Algorithm Based on Information Entropy and Structural Characteristics
Fig. 1. Schematic of Criminisi algorithm principle
Fig. 2. Schematic of difference in block structure to be repaired. (a) Block A to be repaired; (b) block B to be repaired
Fig. 3. Criminisi algorithm inpainting results. (a) Original image; (b) broken image; (c) Criminisi algorithm; (d) partial enlargement image
Fig. 4. Comparison of inpainting results before and after algorithm improvement. (a) Original image; (b) broken image; (c) Criminisi algorithm; (d) proposed algorithm
Fig. 5. Curves of relationship between weight parameters and average PSNR
Fig. 6. Comparison of effects of five algorithms on inpainting artificially broken murals. (a) Original images; (b) artificially adding broken images; (c) Criminisi algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [17]; (f) algorithm in Ref. [18]; (g) proposed algorithm
Fig. 7. Comparison of effects of five algorithms on inpainting real damage murals. (a) Original damage murals; (b) adding mask images; (c) Criminisi algorithm; (d) algorithm in Ref. [12]; (e) algorithm in Ref. [17]; (f) algorithm in Ref. [18]; (g) proposed algorithm
|
Get Citation
Copy Citation Text
Yong Chen, Yapeng Ai, Jin Chen. Dunhuang Mural Inpainting Algorithm Based on Information Entropy and Structural Characteristics[J]. Laser & Optoelectronics Progress, 2020, 57(12): 121020
Category: Image Processing
Received: Nov. 15, 2019
Accepted: Dec. 6, 2019
Published Online: Jun. 3, 2020
The Author Email: Chen Yong (edukeylab@126.com)